Alliances : The Case of Wisconsin Welfare Politics , 1993 – 99

Interest group scholars have long emphasized the importance of group alliances in the policymaking process. But little is known about how groups choose specific alliance partners; that is, who works with whom? Social embeddedness theory suggests that the social location of groups in issue networks affects the information available to them about potential partners and the desirability of particular alliances. To test this hypothesis, I use data from interviews with representatives of 57 interest groups and 46 other significant political actors involved in Wisconsin’s 1993–99 welfare policy debate to model alliance formation with two-stage conditional maximum likelihood regression (2SCML). I find substantial support for my social embeddedness hypotheses that alliance formation is encouraged by previous network interaction, contact with mutual third parties, and having a central position in a network. In short, the placement of groups in networks serves to facilitate alliances among some pairs of groups and to cut off potential connections among others. The last quarter century has witnessed an explosion of interest group formation and participation in policymaking at the state and national levels in the United States (Baumgartner and Jones 1993; Berry 1999; Gray and Lowery 1996). As interest group communities have become larger and more complex, individual groups may find themselves to be less influential than in the past (Salisbury 1990). This crowding of interest communities has led groups to devise new strategies to gain policymaking influence. Forming alliances with other groups is one such strategy. Groups that once tended to operate independently now find that a scarcity of resources and increased competition leads them to seek out alliances (Baumgartner and Jones 1993; Gray and Lowery 1998). Alliances allow groups to demonstrate better the